Latent Relation Representations for Universal Schemas

TL;DR

This paper introduces a universal schema approach for relation extraction that combines surface form predicates and structured database relations, enabling better reasoning over unstructured and structured data with improved accuracy.

Contribution

It proposes matrix factorization models for universal schemas, outperforming traditional classification and state-of-the-art distant supervision methods.

Findings

Higher accuracy than traditional classification methods

Effective integration of unstructured text and structured databases

Outperforms existing distant supervision systems

Abstract

Traditional relation extraction predicts relations within some fixed and finite target schema. Machine learning approaches to this task require either manual annotation or, in the case of distant supervision, existing structured sources of the same schema. The need for existing datasets can be avoided by using a universal schema: the union of all involved schemas (surface form predicates as in OpenIE, and relations in the schemas of pre-existing databases). This schema has an almost unlimited set of relations (due to surface forms), and supports integration with existing structured data (through the relation types of existing databases). To populate a database of such schema we present a family of matrix factorization models that predict affinity between database tuples and relations. We show that this achieves substantially higher accuracy than the traditional classification approach. More importantly, by operating simultaneously on relations observed in text and in pre-existing structured DBs such as Freebase, we are able to reason about unstructured and structured data in mutually-supporting ways. By doing so our approach outperforms state-of-the-art distant supervision systems.